The Suppressive Effect of Enamel MatrixDerivative on Osteocalcin GeneExpression of Osteoblasts Is Neutralizedby an Antibody Against TGF-bYoshiyuki Wada,* Hidekazu Yamamoto,* Satoshi Nanbu,* Morimichi Mizuno,*and Masato Tamura*

Background: Enamel matrix derivative (EMD) plays a cru-cial role in periodontal tissue regeneration. However, the pre-cise mechanism of tissue regeneration by EMD remainsobscure. The purpose of this study was to clarify what factorspresent in EMD show bioactivity.

Methods: We first examined the effect of EMD on MC3T3-E1 osteoblastic cells. To evaluate the differentiation, the ex-pression of osteoblast-related genes was measured by reversetranscription-polymerase chain reaction, and the osteocalcin(OCN) content was measured by enzyme-linked immunosor-bent assay. Alkaline phosphatase activity and the mineral-ization were examined histologically. EMD (intact EMD) wasfiltrated to separate the soluble fraction (soluble EMD), andthe effects of soluble and intact EMD were examined. Neutral-ization of the bioactivity of EMD was performed using a poly-clonal antibody against porcine transforming growth factor-beta(TGF-b).

Results: EMD inhibited the expression of osteoblastic phe-notypes, and we used the inhibitory effect of EMD on osteo-blastic differentiation as a benchmark of activity of EMD.The soluble fraction separated from EMD inhibited osteoblast-related gene expression and OCN synthesis. Soluble EMDsuppressed the OCN gene level within 24 hours, and the effectof soluble EMD mimicked that of TGF-b (10 ng/ml). The anti-body against TGF-b diminished the inhibitory effect of solubleEMD on OCN gene expression.

Conclusions: The inhibitory effect of EMD on OCN gene ex-pression of osteoblastic cells is neutralized by the antibodyagainst TGF-b in it. This result might indicate that EMD con-tains TGF-b and that it participates in the bioactivity of EMD.J Periodontol 2008;79:341-347.

KEY WORDS

Differentiation; neutralization tests; osteoblast; osteocalcin;transforming growth factor-beta.

Several attempts have been madeto restore periodontal tissue lossdue to periodontitis. Enamel ma-

trix derivative (EMD), which is extractedfrom developing porcine tooth germs,has been used widely as an effectivematerial to regenerate periodontal tis-sues.1 The clinical use of EMD is basedon the observation that Hertwig’s epi-thelial root sheath deposits an enamel-like matrix on the dentin surface of thedeveloping root.2

Furthermore, EMD was reported to in-duce the proliferation and differentiationof periodontal ligament cells,3 epithelialcells, fibroblasts,4 cementoblasts,5 andosteoblasts.5,6 These findings indicatethat EMD regulates the expression ofosteoblast-related genes, alkaline phos-phatase(ALP)activity,mineralization,pro-tein synthesis, and the proliferation ofmost kinds of cells. Although EMD showsclinical efficacy, the precise mechanismof tissue regeneration and which factorsinfluence cell functions are still unclear.

Amelogenin,7,8 which is a typicalenamel matrix protein, is one of the can-didates for factors in EMD that show bio-activity. Amelogenin has been focusedon because of its effects on osteogenicgene expression9 and cell adhesion.10

* Department of Oral Health Science, Graduate School of Dentistry, Hokkaido University,Sapporo, Japan.

doi: 10.1902/jop.2008.070197

J Periodontol • February 2008

341

The other candidates are growth factors, suchas transforming growth factor-beta (TGF-b), andbone morphogenetic proteins (BMPs). It is known thatgrowth factors influence cell activity and that toothgerms synthesize and store growth factors.11,12 Itwas reported that TGF-b was expressed11 and syn-thesized in the tooth germ.12 It also is assumed thatthere is an undetectable amount of growth factor inEMD,13,14 although growth factors have not been de-tected in it.15 Furthermore, it was reported that EMDcontains a TGF-b–like factor or BMP-like factorand stimulates the signal transduction of BMP andTGF-b.16 These findings indicate that the effects ofEMD are associated closely with growth factors.

It is not clear which growth factors are present inEMD. To clarify this problem, we used the inhibitoryeffect of EMD on osteoblastic differentiation as abenchmark of its bioactivity. We hypothesized thatthe effects of EMD on cells were dependent on growthfactors, such as TGF-b.

In this study, we found that EMD markedly inhibitedOCN gene expression, OCN protein synthesis, andmineralization. The effect of the soluble fraction ofEMD mimicked that of a high concentration of TGF-b.Furthermore, the inhibitory effect of soluble EMD onOCN gene expression was diminished by an antibodyagainst TGF-b.

Therefore, we concluded that the inhibitory effect ofEMD on osteoblastic differentiation was due to TGF-b

in it and speculated that TGF-b contained in EMDtakes part in its bioactivity.

MATERIALS AND METHODS

Material PreparationEMD was purchased as a gel.† Intact EMD and the sol-uble fraction of EMD (soluble EMD) were prepared be-fore use. Intact EMD was added to the culture mediumat a concentration of 100 mg/ml. After stirring for 48hours, it was filtrated‡ (diameter: 0.2 mm) as solubleEMD. The aggregates were observed on the filter afterthe titration of EMD. Porcine TGF-b§ was recon-structed in sterile 4 mM HCl containing 0.1% bovineserum albumin to prepare a stock solution beforeuse. The polyclonal antibodyi against porcine TGF-b

was used for the neutralization experiment.

Sodium Dodecyl Sulfate–Polyacrylamide SlabGel ElectrophoresisIntact EMD was mixed with phosphate buffered salineand centrifuged to remove the aggregates. The super-natant was filtrated (diameter 0.2 mm) as the solublefraction of EMD (soluble EMD). Intact EMD and thesoluble fraction of EMD were lyophilized.

Next, intact EMD and the soluble fraction ofEMD were subjected to sodium dodecyl sulfate–polyacrylamide slab gel (15%) electrophoresis to

analyze the homogeneity of the material and the mo-lecular weights. The gels were stained with Coomassiebrilliant blue.

Cell CultureOsteoblastic MC3T3-E1 cells that originated frommouse calvaria were cultured in a-minimum essentialmedium¶ that contained ascorbic acid (50 mg/ml)with 10% fetal calf serum.17 Then the cells were platedin a 35-mm dish# and cultured in a humidified atmos-phere of 5% CO2 (volume/volume) at 37�C. The me-dium was changed every 2 days.

Cells were cultured with 100 mg/ml EMD in a long-term experiment. In another experiment, cells werecultured with intact EMD or soluble EMD. The mate-rials were added at the time of dispersal.

In a short-term experiment, cells were cultured withsoluble EMD or 10, 1, or 0.1 ng/ml porcine TGF-b.The materials were added 2 weeks after the cellsreached confluence.

Enzyme-Linked Immunosorbent AssayThe OCN content in the medium was examined usingan enzyme-linked immunosorbent assay (ELISA)commercial kit.** The culture medium was used forthe analysis of OCN content.

Histologic Observation of ALP Activityand MineralizationThe culture of osteoblastic cells was fixed with para-formaldehyde and then washed with distilled waterseveral times. For the detection of mineralization,the culture was stained with 0.1% alizarin red S. Forthe detection of ALP activity, a solution†† dissolvedin 0.1 M Tris buffer (pH 9.3) was poured into thedishes.

Reverse Transcription-PolymeraseChain Reaction (RT-PCR)The mRNA levels of GAPDH, OCN, osteopontin(OPN), and runt-related gene 2 (Runx2) were mea-sured by RT-PCR. The gene-specific primers usedwere: forward CACCATGGAGAAGGCCGGGG andreverse GACGGACACATTGGGGGTAG for GAPDH;forward CGCTACCTTGGAGCCTCAG and reverseAGCCGAGCTGCCAGAGTTTG for OCN; forwardGCTTCTGAGCATGCCCTCTGATCACGA and reverseGTGCAGAAGCTTTTGGTTACAACGGT for OPN; andforward GTGCAGAAGCTTTTGGTTACAACGGT andreverse CGCTCCGGCCCACAAATCTC for Runx2.

† Emdogain, Seikagaku, Tokyo, Japan.‡ Millipore filter, Sartorius Biotech Japan, Tokyo, Japan.§ R&D Systems, Minneapolis, MN.i Abcam, Cambridge, U.K.¶ GIBCO-BRL Life Technologies, Rockville, MD.# Corning International, Tokyo, Japan.** Biomedical Technologies, Stoughton, MA.†† 0.25% naphthol AS-BI phosphate and 0.75% Fast Blue BB, Sigma-

Aldrich, Tokyo, Japan.

Antibody Against TGF-b Neutralizes the Effect of EMD on Osteoblasts Volume 79 • Number 2

342

Total RNA was isolated using the acid guanidinethiocyanate-phenol-chloroform method. The mRNAswere converted to cDNA by avian myeloblastosisvirus–reverse transcriptase.‡‡ The reaction mixtureconsisted of PCR buffer, 2.5 mM MgCl2, 1 mM deoxy-nucleotide triphosphate mixture, 0.2 mM of eachprimer, and Taq polymerase§§ (2.5 U/100 ml). PCRamplification of cDNA was conducted using gene-specific primers in a PCR thermocycler. Thermal cy-cling was performed with one cycle of 94�C for 12minutes, and 30 cycles of 95�C for 30 seconds,60�C for 30 seconds, and 72�C for 90 seconds. Thereaction products were analyzed by electrophoresisof 10-ml samples in 2.5% agarose gels. The amplifiedDNA fragments were stained with ethidium bromideand photographed under ultraviolet illumination.GAPDH mRNA was used as an internal control. Thedensities of the bands were converted to relative val-ues,ii and each gene level was normalized by dividingit by the amount of GAPDH mRNA.

Neutralization of TGF-bThe polyclonal antibody against porcine TGF-b (100ng/ml) was added to soluble EMD and incubated for24 hours. Cells were cultured for 72 hours in the me-dium containing the materials. The porcine TGF-b

was used as the positive control.

Statistical AnalysisAssays were run in triplicate, and the data were ana-lyzed using the Student t test. The experimental andcontrol groups were compared.

RESULTS

Effect of EMD on the Expression ofOsteoblastic PhenotypesThe MC3T3-E1 osteoblastic cells were cultured in amedium containing 100 mg/ml EMD for 3 weeks. Ac-cording to previous reports,5,6 this concentration ofEMD is effective for osteoblastic proliferation and dif-ferentiation. By the culture of osteoblasts with EMD,OCN mRNA expression and protein synthesis weresuppressed markedly for 1 to 3 weeks (Figs. 1A and1B). Furthermore, EMD markedly inhibited minerali-zation and ALP activity at 2 and 3 weeks (Fig. 1C).

These results indicated that EMD suppressed theexpression of the osteoblastic phenotype, and weused this suppressive effect of EMD as a benchmarkof bioactivity of EMD for investigating what factorsin EMD had bioactivity.

Electrophoretic Profiles of Intact EMD and SolubleFraction of EMDThe profiles of low molecular bands in the solubleEMD lane were similar to those in the intact EMD lane(Fig. 2A).

Effect of Soluble Factor Separated From EMDon Osteoblast-Related Gene Expressionand OCN SynthesisWe investigated whether soluble EMD affected otherosteoblast-related gene levels. At 2 weeks of culture,OCN and Runx2 mRNA levels were suppressed by in-tact and soluble EMD at 72 hours after their addition,although the OPN mRNA level was not changed byEMD. These results indicated that the soluble fractionof EMD contained a factor that affected osteoblasts(Fig. 2B).

Effect of Soluble EMD on OCN mRNA Level in aShort-Term ExperimentCells were cultured with the soluble factor for 24hours, and we found that it decreased the OCN mRNAlevel rapidly at 12 hours. The inhibitory effect of sol-uble EMD was not observed at 3 and 6 hours (Fig.3). When soluble fractions of EMD were added, themean expression levels of OCN were 0.54-, 0.56-,1.01-, and 1.15-fold of the control at 24, 12, 6, and3 hours, respectively.

These results showed that the factor present inEMD was water soluble, and its effect was observedwithin a short time. We hypothesized that the solublefactor present in EMD might be a growth factor.

Similarity of Porcine TGF-b and Soluble EMD onOCN mRNA ExpressionNext, we investigated the similarity of soluble EMDand TGF-b. We added 0.1, 1, or 10 ng/ml porcineTGF-b to the medium at 2 weeks after cells reachedconfluence, and the total RNA was isolated at 24 hoursafter adding the materials. TGF-b inhibited OCN geneexpression, and 10 ng/ml TGF-b was the most effec-tive concentration (Fig. 4).

Neutralization of the Suppressive Effect of EMDby the Antibody Against TGF-bCells were cultured for 72 hours in the soluble EMDcontaining medium previously treated with the poly-clonal antibody against TGF-b at 2 weeks after cellsreached confluence. Although the antibody againstTGF-b itself did not influence the OCN mRNA expres-sion of cells, the inhibitory effect of soluble EMD on theOCN mRNA expression was diminished by the treat-ment with the antibody against TGF-b (Fig. 5).

DISCUSSION

In this study, we demonstrated that EMD suppressedthe OCN mRNA level of MC3T3-E1 osteoblastic cells.OCN is known as a typical osteoblastic phenotypeand is associated with mineralization.18

‡‡ Promega, Tokyo, Japan.§§ Applied Biosystems, Tokyo, Japan.ii NIH Image, National Institutes of Health, Bethesda, MD.

J Periodontol • February 2008 Wada, Yamamoto, Nanbu, Mizuno, Tamura

343

Hagewald et al.,6 using a three-dimensional organ-oid culture, reported that EMD enhanced ALP activityand calcium content. However, it might be possiblethat their culture system itself significantly influencedosteoblastic differentiation. In addition, they usedb-glycerophosphate to boost mineralization. We useda simple culture system using MC3T3-E1 cells with-out b-glycerophosphate. Furthermore, they did notinvestigate OCN expression, the typical marker of os-teoblastic differentiation.

Conversely, our results corresponded to a report byTokiyasu et al.5 with regard to the finding that EMD in-hibits OCN gene expression. Furthermore, we foundthat EMD inhibited OCN protein synthesis, minerali-zation, and ALP activity of osteoblasts. Although100 mg/ml EMD is a relatively high concentration, itis appropriate to investigate the effects of EMD on cellactivity because it is applied directly to the root sur-face. We used this inhibitory effect of EMD on OCNgene expression as a benchmark of its bioactivityfor analyzing its bioactive constituents in this study.

Two candidates showing bioactivity are present inEMD. One consists of growth factors11,12 and theother consists of enamel matrix proteins.7,8 Enamelmatrix proteins are known to aggregate and form aninsoluble matrix.19 Therefore, we separated the solu-ble fraction (soluble EMD) from intact EMD to exclude

the possibility that the insoluble matrix showedbioactivity. We demonstrated that soluble EMD in-hibited the expression of OCN and Runx2 mRNAs.Thus, our results demonstrated that a soluble factorin EMD possessed the ability to affect cells.

He et al.20 reported that direct contact betweenEMD and osteoblasts was not required for EMD-induced cell proliferation; they insisted that solublefactors in EMD affected osteoblastic proliferationand speculated that the factors were growth factorsor small molecules of enamel matrix. Our resultsagreed with their report with regard to the point thatthe crucial factor was present in the soluble fractionof EMD, and we demonstrated that the factor influ-enced cell proliferation and cell differentiation.

Next, we investigated how much time was neces-sary to confirm that growth factors participated inthe inhibitory effect that EMD had on the expressionof osteoblastic phenotypes. Soluble EMD suppressedthe OCN mRNA level within 24 hours after its addition.This feature is common with growth factors, but notwith matrix proteins. This result supported the ideathat a growth factor present in EMD played some rolein the expression of its bioactivity.

TGF-b modulates the growth and differentiation ofvarious cell types.21,22 It is produced by osteoblastsand is stored in bone matrix in a latent form.23 TGF-b

Figure 1.Effect of EMD on phenotypes of osteoblastic differentiation. A) The OCN gene expression clearly was suppressed by EMD (100 mg/ml) for 1 to 3weeks. B) The OCN protein content in the medium was measured using the ELISA method for 1 to 3 weeks. Each column shows the mean – SD(N = 3). At 2 and 3 weeks, the OCN protein content was decreased significantly (*P <0.05) by EMD. C) The mineralization was suppressedmarkedly by EMD at 3 weeks. The ALP activity clearly was suppressed by EMD at 3 weeks. CT = control; w = week(s).

Antibody Against TGF-b Neutralizes the Effect of EMD on Osteoblasts Volume 79 • Number 2

344

prevents the expression of ALP and OCN genes inosteoblastic differentiation.24 It was reported thatthe effect of TGF-b on the expression of osteoblasticphenotypes was suppressed when osteoblastic cellsdifferentiated into the mature stage25 and that itsuppressed ALP activity in MC3T3-E1 osteoblasticcells.26 Furthermore, Alliston et al.27 reported thatTGF-b inhibited OCN mRNA expression by inhibitionof Runx2 via Smad downstream. These reports con-cerning TGF-b are consistent with our finding thatEMD inhibited the Runx2 mRNA level and the OCNmRNA level.

Using a radioimmunoassay, Gestrelius et al.15 re-ported that EMD obtained from porcine tooth germdid not contain growth factors. However, TGF-b

mRNA was expressed in tooth germ,11 and TGF-b

was detected as a protein in it.12 Recently, it was re-ported that EMD showed TGF-b–like activity,13 andSuzuki et al.16 reported that EMD stimulated signaltransduction of TGF-b. These reports strongly sug-gest that TGF-b is present in EMD, and we considerthat the result of Gestrelius et al.15 was not conclusive.

To support our hypothesis that TGF-b present inEMD suppressed the OCN gene expression of osteo-blasts, we compared the effects of soluble EMD andporcine TGF-b on OCN mRNA expression and foundthat both inhibited OCN gene expression. Althoughthe TGF-b concentration of 10 ng/ml was high com-pared to that present in tissue, other soluble factors,e.g., growth factors other than TGF-b or solubleenamel matrix proteins, might be involved in the effectof EMD.

To verify the possibility that the suppressive factorof EMD on OCN gene expression of osteoblasts wasTGF-b, we examined whether an antibody againstTGF-b could abolish the effect of EMD. We found thatthe inhibitory effect of EMD was diminished markedlyby the antibody. This result strongly suggested thatEMD affected the expression of osteoblastic pheno-types via TGF-b. Therefore, we concluded EMD mightcontain TGF-b that exerts an effect on osteoblasts.

Figure 2.Electrophoretic profiles of intact EMD and soluble fraction of EMD.A) The profiles of low molecular bands in the soluble EMD lane weresimilar to those in the intact EMD lane. Effects of intact EMD andsoluble fraction of EMD (soluble EMD) on osteoblast-related geneexpression. B) RT-PCR was performed to examine the effects of intactEMD and soluble EMD on osteoblast-related gene expression at2 weeks after cells reached confluence. The OCN and Runx2 mRNAlevels were clearly suppressed by intact EMD and soluble EMD. TheOPN mRNA level was not influenced. CT = control; STD = standard;CBB = Coomassie brilliant blue.

Figure 3.Effect of soluble EMD on the OCN mRNA level in a short-termexperiment. CT = control; h = hours.

Figure 4.Comparison between the effects of TGF-b and soluble EMD on theOCN mRNA level. The inhibitory effects of soluble EMD and 10 ng/mlTGF-b were comparable. CT = control.

Figure 5.Effect of antibody against TGF-b on the inhibitory effect of solubleEMD on OCN gene expression. The inhibitory effect of soluble EMDon OCN mRNA expression was diminished by the antibody againstTGF-b. Porcine TGF-b (10 ng/ml) was used as a positive control.CT = control.

J Periodontol • February 2008 Wada, Yamamoto, Nanbu, Mizuno, Tamura

345

The usefulness of EMD has been demonstratedwidely. However, our results indicated that EMD hasan inhibitory effect on osteoblastic differentiation.EMD might have different effects, depending on thesource of cells and cell condition, e.g., proliferationstage or differentiation stage. If the positive effectsof EMD are mediated by TGF-b, the effect of EMDon periodontal tissue regeneration might be duemainly to its effect on cell proliferation. Simulta-neously, TGF-b might regulate cell differentiationproperly for tissue regeneration. Therefore, the timingof the application of EMD in the wound-healing stageand its dosage could be crucial factors for successfulperiodontal therapy because the growth factors haveeffects that depend on their dose or cell condition.21

CONCLUSIONS

EMD is composed of several substances. One of themain constituents is amelogenin. It aggregates andforms an insoluble matrix that promotes cell adhesionand cell spreading.10 Amelogenin stimulates the bonesialoprotein (BSP) expression of rat osteoblast-likeosteosarcoma 17/2.8 cells through the fibroblastgrowth factor 2 response element and the TGF-b1 ac-tivation element in the promoter of the BSP gene.28

These findings indicate that amelogenin possessessome bioactivity; however, the precise mechanismof action of amelogenin on cells remains obscure.Our results do not exclude the possibility that thesmall molecule of amelogenin affects cell functions.Amelogenin is considered a scaffold of growth factorsthat may regulate their release, including that of TGF-b. Moreover, EMD may contain several growth fac-tors, such as BMP; therefore, further study is neces-sary to estimate their functions.

ACKNOWLEDGMENTS

This work was supported by a grant-in-aid from theMinistry of Education, Culture, Sports, Science andTechnology of Japan. The authors report no conflictsof interest related to this study.

REFERENCES1. Venezia E, Goldstein M, Boyan BD, Schwartz Z. The use

of enamel matrix derivative in the treatment of peri-odontal defect: A literature review and meta-analysis.Crit Rev Oral Biol Med 2004;15:382-402.

2. Hammarstrom L. Enamel matrix, cementum develop-ment and regeneration. J Clin Periodontol 1997;24:658-668.

3. Takayanagi K, Osawa G, Nakaya H, Cochran DL,Kamoi K, Oates TW. Effects of enamel matrix deriv-ative on bone-related mRNA expression in human peri-odontal ligament cells in vitro. J Periodontol 2006;77:891-898.

4. Haase HR, Bartold PM. Enamel matrix derivativeinduces matrix synthesis by cultured human periodon-tal fibroblast cells. J Periodontol 2001;72:341-348.

5. Tokiyasu Y, Takata T, Saygin E, Somerman M. Enamelfactors regulate expression of genes associated withcementoblasts. J Periodontol 2000;71:1829-1839.

6. Hagewald S, Pischon N, Jawor P, Bernimoulin JP,Zimmermann B. Effects of enamel matrix derivative onproliferation and differentiation of primary osteoblasts.Oral Surg Oral Med Oral Pathol Oral Radiol Endod2004;98:243-249.

7. Fincham AG, Moradian-Oldak J. Recent advances inamelogenin biochemistry. Connect Tissue Res 1995;32:119-124.

8. Yamakoshi Y, Tanabe T, Fukae M, Shimizu M. Porcineamelogenins. Calcif Tissue Int 1994;54:277-281.

9. Du C, Schneider R, Zaharias R, et al. Apatite/amelo-genin coating on titanium promotes osteogenic geneexpression. J Dent Res 2005;84:1070-1074.

10. Hoang AM, Klebe RJ, Steffensen B, Ryu OH, SimmerJP, Cochran DL. Amelogenin is a cell adhesion pro-tein. J Dent Res 2002;81:497-500.

11. Jepsen S, Schiltz P, Strong DD, Scharla SH, Snead ML,Finkelman RD. Transforming growth factor-beta1 mRNA in neonatal ovine molars visualized by in situhybridization: Potential role for the stratum interme-dium. Arch Oral Biol 1992;37:645-653.

12. D’Souza RN, Happonen RP, Ritter NM, Butler WT.Temporal and spatial patterns of transforming growthfactor-beta 1 expression in developing rat molars.Arch Oral Biol 1990;35:957-965.

13. Kawase T, Okuda K, Momose M, Kato Y, Yoshie H,Burns DM. Enamel matrix derivative (EMDOGAIN)rapidly stimulates phosphorylation of the MAP kinasefamily and nuclear accumulation of smad2 in both oralepithelial and fibroblastic human cells. J PeriodontalRes 2001;36:367-376.

14. Kawase T, Okuda K, Yoshie H, Burns DM. Anti-TGF-bantibodyblocksenamelmatrixderivative-inducedupreg-ulation of p21WAF1/cip1 and prevents its inhibition ofhuman oral epithelial cell proliferation. J Periodontal Res2002;37:255-262.

15. Gestrelius S, Andersson C, Lidstrom D, HammarstramL, Sommerman M. In vitro studies on periodontal liga-ment cells and enamel matrix derivative. J Clin Peri-odontol 1997;24:685-692.

16. Suzuki S, Nagano T, Yamakoshi Y, et al. Enamelmatrix derivative gel stimulates signal transduction ofBMP and TGF-beta. J Dent Res 2005;84:510-514.

17. Kodama H, Amagai Y, Sudo H, Kasai S, Yamamoto S.Establishment of a clonal osteogenic cell line fromnewborn mouse calvaria. Jpn J Oral Biol 1981;121:899-901.

18. Price PA, Otsuka AA, Poser JW, Kristaponis J, Raman N.Characterization of a g-carboxyglutamic acid-containingprotein from bone. Proc Natl Acad Sci USA 1976;73:1447-1451.

19. Tan J, Leung W, Moradian-Oldak J, Zeichner-DavidM, Fincham AG. Quantitative analysis of amelogeninsolubility. J Dent Res 1998;77:1388-1396.

20. He J, Jiang J, Safavi KE, Spangberg LSW, Zhu Q. Directcontact between enamel matrix derivative (EMD) andosteoblasts is not required for EMD-induced cell prolif-eration. Oral Surg Oral Med Oral Pathol Oral RadiolEndod 2004;98:370-375.

21. Sporn MB, Roberts AB, Wakefield LM, Assoian RK.Transforming growth factor-beta: Biological functionand chemical structure. Science 1986;233:532-534.

22. Masui T, Wakefield LM, Lechner JF, LaVeck MA,Sporn MB, Harris CC. Type beta transforming growth

Antibody Against TGF-b Neutralizes the Effect of EMD on Osteoblasts Volume 79 • Number 2

346

factor is the primary differentiation-inducing serumfactor for normal human bronchial epithelial cells. ProcNatl Acad Sci USA 1986;83:2438-2442.

23. Centrella M, Canalis E. Transforming and nontransform-ing growth factors are present in medium conditioned byfetal calvaliae. Proc Natl Acad Sci USA 1985;82:7335-7339.

24. Ibbotson KJ, Orcutt CM, Anglin AM, D’Souza SM.Effects of transforming growth factor beta 2 on amouse clonal, osteoblastlike cell line MC3T3-E1.J Bone Miner Res 1989;4:37-45.

25. Takeuchi Y, Nakayama K, Matsumoto T. Differentia-tion and cell surface expression of transforminggrowth factor-beta receptors are regulated by interac-tion with matrix collagen in murine osteoblastic cells.J Biol Chem 1996;271:3938-3944.

26. Mizuno M, Fujisawa R, Kuboki Y. Carboxyl-terminalpropeptide of type I collagen (c-propeptide) modu-lates the action of TGF-b on MC3T3-E1 osteoblasticcells. FEBS Lett 2000;479:123-126.

27. Alliston T, Choy L, Ducy P, Karsenty G, Derynck R.TGF-b-induced repression of CBFA1 by Smad3 de-creases cbfa1 and osteocalcin expression and inhibitsosteoblast differentiation. EMBO J 2001;20:2254-2272.

28. Shimizu E, Saito R, Nakayama Y, et al. Amelogeninstimulates bone sialoprotein (BSP) expression throughfibroblast growth factor 2 response element and trans-forming growth factor-beta1 activation element in thepromoter of the BSP gene. J Periodontol 2005;76:1482-1489.

Correspondence: Dr. Yoshiyuki Wada, Department of OralHealth Science, Graduate School of Dentistry, HokkaidoUniversity, Nishi 7, Kita 13, Kita-ku, Sapporo 060-8586,Japan. Fax: 81-11-706-4235; e-mail: wada4888@jeans.ocn.ne.jp.

Submitted April 11, 2007; accepted for publication July18, 2007.

J Periodontol • February 2008 Wada, Yamamoto, Nanbu, Mizuno, Tamura

347